Governance of competition in the Swiss and European railway sector

(1)

Governance of competition in the Swiss and European

railway sector

Final research report to the SBB lab, University of St.Gallen May 2012

“Rail companies’ fiercest competition comes from other transport modes. Efficient competition between modes of transport depends on achieving appropriate pricing and

investment policies.” (Nash and Matthews, 2009)

Authors

Prof Matthias Finger, Ecole Polytechnique Fédérale de Lausanne and European University Institute

Dr Andrea Rosa, European University Institute With the collaboration of

Dr Marc Laperrouza, Ecole Polytechnique Fédérale de Lausanne Dr Martin Holterman, European University Institute

Ian Brand-Weiner, European University Institute Nadia Bert, European University Institute

(2)

Robert Schuman Centre for Advanced Studies

Within the European University Institute (EUI), the Robert Schuman Centre for Advanced Studies (RSCAS), created in 1992 and directed by Stefano Bartolini since September 2006, aims to develop inter-disciplinary and comparative research and to promote work on the major issues facing the process of integration and European society.

The Centre is home to a large post-doctoral programme and hosts major research programmes and projects, and a range of working groups and ad hoc initiatives. The research agenda is organised around a set of core themes and is continuously evolving, reflecting the changing agenda of European integration and the expanding membership of the European Union.

Details of the research of the Centre can be found on: http://www.eui.eu/RSCAS/Research/

Research publications take the form of Working Papers, Policy Papers, Distinguished Lectures and books. Most of these are also available on the RSCAS website:

http://www.eui.eu/RSCAS/Publications/

The EUI and the RSCAS are not responsible for the opinion expressed by the author(s).

Florence School of Regulation

The Florence School of Regulation (FSR) is a partnership between the Robert Schuman Centre for Advanced Studies (RSCAS) at the European University Institute (EUI), the Council of the European Energy Regulators (CEER) and the Independent Regulators Group (IRG). Moreover, as part of the EUI, the FSR works closely with the European Commission. The objectives of the FSR are to promote informed discussions on key policy issues, through workshops and seminars, to provide state-of-the-art training for practitioners (from European Commission, National Regulators and private companies), to produce analytical and empirical researches about regulated sectors, to network, and to exchange documents and ideas.

At present, its scope is focused on the regulation of Energy (electricity and gas markets), of Communications & Media, and of Transport.

This research project has been carried out within the scope of the research activities of FSR - Transport Area.

For further information

Florence School of Regulation – Transport Area Robert Schuman Centre for Advanced Studies European University Institute

Via Boccaccio, 151 I-50133 Firenze

Tel.: +39 055 4685 751 Fax: +39 055 4685 755

Transport Area E-mail: fsr.transport@eui.eu http://fsr.eui.eu

(3)

Summary

This report looks at the railway systems in selected European countries (France, Germany, Great Britain, the Netherlands, Sweden, and Switzerland) with a focus on passenger services with the aim to understand whether the institutional approaches to governing those systems may be typified and whether the approaches or the types may be related to some measure of systemic performance. Overall, this research aims at drawing lessons of general use from the situations examined.

This research is based on three sets of information. First, an in-depth account of the current state of railway systems in the countries examined is provided. This has been carried out with a particular focus on regulatory arrangements, Public Service Obligations (PSOs) and overall evolution of railway-related institutions over the years since 1988. The main purpose of such a portrait is to observe and characterise the types of institutional arrangements. Then publicly available performance indicators (PIs) have been collected and contrasted with the institutional evolutions, seeking a causal link. Finally, a set of interviews with stakeholders has been carried out in all countries except Switzerland to investigate the existence of a systemic approach, as well as the stakeholders’ understanding of performance.

The analysis leads us to conclude that each country is a type: there is a variety of institutional approaches and, in terms of institutional arrangements, convergence among countries cannot be detected. While the policies and different Directives of the European Commission certainly do play a role, the institutional path of each country is essential in respectively defining the institutional approaches. Moreover, no country seems to be entirely settled into a lasting system, even when developments have been introduced in stages. The need for further evolution seems particularly important when it comes to interfaces among actors. We have observed, however, a certain movement towards regionalisation, which may be taken as an emerging trend with so far satisfactory results. We have also observed the importance of PSO services for new entrants and the link of such services to the regionalisation trend. Open access is of minor relevance in all the States surveyed.

The establishment of a relationship between the evolution of the institutional arrangements and of the publicly available systemic indicators has been unattainable. At times, improvements in performance that coincide with institutional reforms were perceived. However, such instances are much less common than they would have to be in order to count as clear evidence of the effect of the reforms in question. This can mean that either performance evolves irrespective of the institutional arrangements or the institutional arrangements produce effects only with a time lag beyond the scope of this research. Other possible explanations are that the PIs commonly used do not significantly respond to the change in institutional arrangements, given that the institutional arrangements are aimed at producing systemic effects whereas most of the measured PIs respond to firm behaviour. The interviews led us to conclude that no one speaks for the railway system of a country. A system view is lacking while there is a demand for a body with a systemic function. Depending on the system, such a body may be the incumbent integrated operator, the Infrastructure Manager (IM) at arm’s length of the Ministry, or the Regulator. The interviews looked also at system PIs, and disagreement as to what good system-wide indicators are emerged. The difficulty in relating systemic indicators to the system’s evolution may be due to the lack of system’s PIs. Also, some indicators mentioned have less to do with the output of the railway system, but rather with the outcomes of a given policy or with input variables.

(4)

Our interviews also show that each actor defines its own PIs, which are most likely the ones that suit the actor better than the system. These stakeholder PIs are also not harmonized, nor integrated nor consolidated, thus leading to a multiplication of measurements and ultimately to the impossibility of defining performance in a coherent manner.

(5)

List of figures

Figure 1. Main actors of French railway system ... 37 Figure 2. Evolution of institutions in the French railway sector, 1988-2011. The full size picture is too big to be included here and is attached as a spreadsheet to this report. The overview has been obtained by collecting information from the literature ... 45 Figure 3. Milestones in the evolution of institutions in the French rail sector, 1988-2011. The full size picture is too big to be included here and is attached as a spreadsheet to this report. The overview has been obtained by collecting information from the literature ... 46 Figure 4. Main actors of German railway system ... 48 Figure 5. Evolution of institutions in the German railway sector, 1988-2011. The full size picture is too large to be included here and is attached as a spreadsheet to this report. The overview has been obtained by collecting information from the literature ... 57 Figure 6. Milestones in the evolution of institutions in the German railway sector, 1988-2011. The full size picture is too large to be included here and is attached as a spreadsheet to this report. The overview has been obtained by collecting information from the literature ... 58 Figure 7. Main actors of the British railway system ... 61 Figure 8. Evolution of institutions in the British railway sector, 1988-2011. The full size picture is too large to be included here and is attached as a spreadsheet to this report. The overview has been obtained by collecting information from the literature and, partly, thanks to personal communications of experts... 69 Figure 9. Milestones in the evolution of the railway sector in Great Britain, 1988-2011. The full size picture is too large to be included here and is attached as a spreadsheet to this report. The overview has been obtained by collecting information from the literature and, partly, thanks to personal communications of experts ... 70 Figure 10. Main actors of the Dutch railway system ... 73 Figure 11. Evolution of institutions in the Dutch railway sector, 1988-2011. The full size picture is too big to be included here and is attached as a spreadsheet to this report. The overview has been obtained by collecting information from the literature ... 82 Figure 12. Milestones in the evolution of institutions in the Dutch railway sector, 1988-2011. The full size picture is too big to be included here and is attached as a spreadsheet to this report. The overview has been obtained by collecting information from the literature ... 83 Figure 13. Main actors of Swedish railway system ... 86 Figure 14. Evolution of institutions in the Swedish railway sector, 1988-2011. The full size picture is too big to be included here and is attached as a spreadsheet to this report. The overview has been obtained by collecting information from the literature ... 97 Figure 15. Milestones in the evolution of institutions in the Swedish railway sector, 1988-2011. The full size picture is too big to be included here and is attached as a spreadsheet to this report. The overview has been obtained by collecting information from the literature .... 98 Figure 16. Main actors of Swiss railway system ... 101 Figure 17. Evolution of institutions in the Swiss railway sector, 1988-2011. The full size picture is too big to be included here and is attached as a spreadsheet to this report. The overview has been obtained by collecting information from the literature ... 109

(10)

Figure 18. Milestones in the evolution of institutions in the Swiss railway sector, 1988-2011. The full size picture is too large to be included here and is attached as a spreadsheet to this

report. The overview has been obtained by collecting information from the literature ... 110

Figure 19. Current (2011) institutional set-up in France, Germany, Britain, the Netherlands, Sweden and Switzerland ... 122

Figure 20. Evolution of the Railway Regulators in the case countries ... 133

Figure 21. InteGRail KPI ... 143

Figure 22. The Ding et al. model ... 143

Figure 23. Length of rail lines in km and intensity of use in train-km/km of lines ... 151

Figure 24. Percentage of passenger train-km on the total and modal split ... 152

Figure 25. Modal split for passenger and freight on rail... 152

Figure 26. Relevance of high speed lines with respect to whole network length and importance of high speed passenger transport over the data for the whole network ... 153

Figure 27. Length of railway lines ... 154

Figure 28. Length of high speed lines ... 154

Figure 29. Expenditure of principal railway enterprises in rolling stock, in euros at 1999 price level ... 155

Figure 30. Expenditure of principal railway enterprises in infrasturcture, in euros at 1999 price level ... 155

Figure 31. Gross investment spending in rail infrastructure, in euros at current prices ... 156

Figure 32. Maintenance expenditures spending in rail infrastructure, in euros at current prices ... 156

Figure 33. Number of employed people in the railway sector ... 157

Figure 34. Number of enterprises in the railway sector ... 157

Figure 35. Length of railway lines relative to national surface area ... 158

Figure 36. Total train kilometers, including passengers, goods and other trains... 159

Figure 37. Passengers train kilometers ... 159

Figure 38. Annual number of victims in an accident involving railways ... 160

Figure 39. Annual number of accident involving railways ... 160

Figure 40. Number of accidents involving the transport of dangerous goods ... 161

Figure 41. Annual average indices for transport prices: passenger transport by railway in Case States and Switzerland. The base year is 2005 ... 161

Figure 42. Passengers on rail, measured in million passenger-km ... 162

Figure 43. Passenger traffic on high speed rail ... 163

Figure 44. Share of high speed rail in passenger traffic ... 163

Figure 45. Normed ratio passenger transport (passenger-km) to GDP, with 2000 as base year ... 164

Figure 46. Freight rail traffic. Plot including data for EU15 and EU27 ... 164

(11)

Figure 48. Normed ratio freigth transport (tonne-km) to GDP, with 2000 as base year ... 165

Figure 49. Rail modal split – passengers. Case States, EU15, EU27 ... 166

Figure 50. Rail modal split – freight. Case States, EU15, EU27 ... 166

Figure 51. CO2 emissions from fuel combustion in rail transport [million t] ... 167

Figure 52. NO2 emissions from fuel combustion in rail transport [million t] ... 167

Figure 53. CH4 emissions from fuel combustion in rail transport [million t] ... 168

Figure 54. France: length of lines (km). The rectangles placed vertically across the chart characterise the milestones of the institutional evolution indicated in the previous chapter . 169 Figure 55. France: total train km (1000 train km). The rectangles placed vertically across the chart characterise the milestones of the institutional evolution indicated in the previous chapter ... 170

Figure 56. France: passenger kilometres (1000 mio pkm). The rectangles placed vertically across the chart characterise the milestones of the institutional evolution indicated in the previous chapter ... 171

Figure 57. Rail Infrastructure Gross Investment Spending (million euro, current prices). The rectangles placed vertically across the chart characterise the milestones of the institutional evolution indicated in the previous chapter ... 172

Figure 58. Germany: total length of lines (km) and length of high speed lines (km). The rectangles placed vertically across the chart characterise the milestones of the institutional evolution indicated in the previous chapter ... 173

Figure 59. Germany: total train kilometres (1000 train km). The rectangles placed vertically across the chart characterise the milestones of the institutional evolution indicated in the previous chapter ... 174

Figure 60. Germany: passenger kilometres (1000 mio pkm). The rectangles placed vertically across the chart characterise the milestones of the institutional evolution indicated in the previous chapter ... 175

Figure 61. Germany: tonne kilometres (1000 mio tkm). The rectangles placed vertically across the chart characterise the milestones of the institutional evolution indicated in the previous chapter ... 175

Figure 62. Germany: % of passenger and freight traffic on rail. The rectangles placed vertically across the chart characterise the milestones of the institutional evolution indicated in the previous chapter ... 176

Figure 63. Great Britain: Rail Infrastructure Gross Investment Spending (million euro, current prices). The rectangles placed vertically across the chart characterise the milestones of the institutional evolution indicated in the previous chapter ... 177

Figure 64. Great Britain: passenger kilometres (1000 mio pkm). The rectangles placed vertically across the chart characterise the milestones of the institutional evolution indicated in the previous chapter ... 178

Figure 65. Great Britain: tonne kilometres (1000 mio tkm). The rectangles placed vertically across the chart characterise the milestones of the institutional evolution indicated in the previous chapter ... 178

Figure 66. Great Britain: % of passenger and freight traffic on rail. The rectangles placed vertically across the chart characterise the milestones of the institutional evolution indicated in the previous chapter ... 179

(12)

Figure 67. The Netherlands: Annual number of victims. The rectangles placed vertically across the chart characterise the milestones of the institutional evolution indicated in the previous chapter ... 180 Figure 68. The Netherlands: passenger kilometres (1000 mio pkm). The rectangles placed vertically across the chart characterise the milestones of the institutional evolution indicated in the previous chapter ... 181 Figure 69. The Netherlands: tonne kilometres (1000 mio tkm). The rectangles placed

vertically across the chart characterise the milestones of the institutional evolution indicated in the previous chapter ... 181 Figure 70. The Netherlands: % of passenger and freight traffic on rail. The rectangles placed vertically across the chart characterise the milestones of the institutional evolution indicated in the previous chapter ... 182 Figure 71. Sweden: passenger kilometres (1000 mio pkm). The rectangles placed vertically across the chart characterise the milestones of the institutional evolution indicated in the previous chapter ... 183 Figure 72. Sweden: % of passenger and freight traffic on rail. The rectangles placed vertically across the chart characterise the milestones of the institutional evolution indicated in the previous chapter ... 184 Figure 73. Sweden: tonne kilometres (1000 mio tkm). The rectangles placed vertically across the chart characterise the milestones of the institutional evolution indicated in the previous chapter ... 185 Figure 74. Switzerland: Rail Infrastructure Gross Investment Spending (million euro, current prices). The rectangles placed vertically across the chart characterise the milestones of the institutional evolution indicated in the previous chapter ... 187 Figure 75. Switzerland: passenger kilometres (1000 million pkm). The rectangles placed vertically across the chart characterise the milestones of the institutional evolution indicated in the previous chapter ... 188 Figure 76. Switzerland: tonne kilometres (1000 million tkm). The rectangles placed vertically across the chart characterise the milestones of the institutional evolution indicated in the previous chapter ... 188 Figure 77. Switzerland: % of passenger and freight traffic on rail. The rectangles placed vertically across the chart characterise the milestones of the institutional evolution indicated in the previous chapter ... 189 Figure 78. Train reliability (public performance measure – percentage of trains on time) ... 254 Figure 79. Passenger rail industry expenditure 1996/97 to 2009/10 ... 254 Figure 80. Industry expenditure per passenger-km (2009/10 prices) ... 255 Figure 81. Comparison of whole-system costs (partly normalised) £/'000 passenger-km .... 255 Figure 82. Indexed trends in performance, passenger-km and costs ... 255 Figure 83. Hierarchy of railway outputs ... 256

(13)

List of tables

Table 1. Research questions and sections of the report addressing them ... 22

Table 2. Market opening in Europe (2008) - market shares ... 30

Table 3. Legal state of competition in Europe ... 31

Table 4. Three institutional models of railways in Europe ... 31

Table 5. Selected statistics, France (2005-2009) ... 36

Table 6. Summary of provisions for access to the French rail market ... 36

Table 7. Regulatory institutions relevant to the French railway market ... 36

Table 8. Main information about the Economic Regulator of the French rail industry... 37

Table 9. Infrastructure management and path allocation in the French railway market ... 37

Table 10. Brief account of the main points on the evolution of the regulatory and competition arrangements ... 43

Table 11. Selected statistics, Germany (2005-2009) ... 47

Table 12. Summary of provisions for access to the German rail market ... 47

Table 13. Regulatory institutions relevant to the German railway market ... 47

Table 14. Main information about the Economic (and Safety) Regulator of the German rail industry ... 48

Table 15. Infrastructure management and path allocation in the German railway market ... 48

Table 16. Competition issues dealt with by BDA in 2008 and 2009 ... 52

Table 18. Selected statistics, Great Britain (2005-2009) ... 60

Table 19. Summary of provisions for access to the British rail market ... 60

Table 20. Regulatory institutions relevant to the British railway market ... 60

Table 21. Main information about the Economic Regulator of the British rail industry ... 60

Table 22. Infrastructure management and path allocation in the British railway market ... 60

Table 23. Competition issues dealt with in 2008 and 2009 ... 63

(14)

Table 25. Selected statistics, The Netherlands (2005-2009)... 71

Table 26. Summary of provisions for access to the Dutch rail market ... 71

Table 27. Regulatory institutions relevant to the Dutch railway market ... 72

Table 28. Main information about the economic (and safety) Regulator of the Dutch rail industry ... 72

Table 29. Infrastructure management and path allocation in the Dutch railway market ... 72

Table 30. Brief account of the main points on the evolution of the Dutch regulatory and competition arrangements ... 80

Table 31. Selected statistics, Sweden (2005-2009) ... 84

Table 32. Summary of provisions for access to the Swedish rail market ... 84

Table 33. Regulatory institutions relevant to the Swedish railway market ... 85

Table 34. Main information about the Economic (and Safety) Regulator of the Swedish rail industry ... 85

Table 35. Infrastructure management and path allocation in the Swedish railway market ... 85

Table 36. Examples of effects on subsidies in competitive tenders ... 92

Table 37. Brief account of the main points on the evolution of the regulatory and competition arrangements in Sweden ... 95

Table 38. Selected statistics, Switzerland (2005-2009) ... 99

Table 39. Summary of provisions for access to the Swiss railways ... 99

Table 40. Regulatory institutions relevant to the Swiss railways ... 99

Table 41. Main information about the Arbitration Commission ... 99

Table 42. Infrastructure management and path allocation in the Swiss railways ... 100

Table 43. Brief account of the main points on the evolution of the regulatory and competition arrangements in Switzerland (evolutions of the Regulatory Agencies in italics) ... 107

Table 44. A summary of some of salient aspect of the institutional set-ups reviewed. (*) of limited significance since only one operator is allowed ... 123

Table 45. Long distance market set-ups and the Authority contracting the passenger services ... 124

Table 46. Regional/local passenger set-up and the Authority contracting the passenger services ... 126

(15)

Table 47. Current arrangements for the provision/procurement of rolling stock and

maintenance for services on concession/franchise ... 126

Table 48. Summary of current ownership situation of incumbents in the case countries ... 127

Table 49. IMs and incumbent rail undertakings, current arrangements ... 128

Table 50. Current arrangements for capacity allocation ... 128

Table 51. Current arrangements for traffic control ... 129

Table 52. Current arrangements for station management ... 129

Table 53. Current modal arrangements of IMs ... 130

Table 54. Current arrangements for the entrance of new operators on the national passenger markets (this table does not refer to international passenger traffic) ... 131

Table 55. Current arrangements of Safety Regulators ... 131

Table 56. Multimarket remit of Railway Economic Regulators ... 133

Table 57. Summary of regulatory arrangements... 134

Table 58. Common performance indicators in rail transport ... 142

Table 59. Input indicators and general statistics available, along with source and description ... 147

Table 60. Table of performance indicators considered, along with their source and description ... 148

Table 61. Classification of the available performance indicators according to the framework by Finger et al. (2010) ... 150

(16)

List of attached spreadsheets

GCSR_France_institutional_evolution_withPIs.xlsx: includes the charts in Figure 2 and Figure 3 as well as the facility to plot the indicators discussed in chapter 5

GCSR_Germany_institutional_evolution_withPIs.xlsx: includes the charts in Figure 5 and Figure 6 as well as the facility to plot the indicators discussed in chapter 5

GCSR_Britain_institutional_evolution_withPIs.xlsx: includes the charts in Figure 8 and Figure 9 as well as the facility to plot the indicators discussed in chapter 5

GCSR_Netherlands_institutional_evolution_withPIs.xlsx: includes the charts in Figure 11 and Figure 12 as well as the facility to plot the indicators discussed in chapter 5

GCSR_Sweden_institutional_evolution_withPIs.xlsx: includes the charts in Figure 14 and Figure 15 as well as the facility to plot the indicators discussed in chapter 5

GCSR_Switzerland_institutional_evolution_withPIs.xlsx: includes the charts in Figure 17 and Figure 18 as well as the facility to plot the indicators discussed in chapter 5

(17)

List of acronyms

In brackets, the Case Country to which the term refers: FR (France), DE (Germany), UK (United Kingdom), SE (Sweden), SWI (Switzerland).

AEG Allgemeines Eisenbahngesetz (DE)

AFITF Agence de financement des infrastructures de transport de France (FR) AFRA Association Française du rail (FR)

AOT Autorités Organisatrices de Transport (FR)

ARAF Autorité de régulation des activités ferroviaires (FR) ASTOC Association of Swedish Train Operating Companies (SE) BAV-FOT Bumdesamt für Verkehr, Federal Office of Transport (SWI) BKA Bundeskartellamt (DE)

BLS Bern-Lötschberg-Simplon (SWI)

BMVBS Bundesministerium für Verkehr, Bau und Stadtentwicklung (DE) BMWi Bundesministerium für Wirtschaft und Technologie (DE)

BNetzA Bundesnetz Agentur (DE) BR British Railways (UK)

BSWAG Bundesschienenwegeausbaugesetz (DE)

BU Business Unit

CPTA County Public Transport Authorities (SE)

CSSPF Conseil supérieur du service public ferroviaire (FR)

DB Deutsche Bahn (DE)

DCF Direction de la Circulation Ferroviaire (FR)

DETEC Federal Department of the Environment, Transport, Energy and Communications (SWI)

DfT Department for Transport (UK) EBA Eisenbahnbundesamt (DE) EdB Eisenbahnen des Bundes (DE)

EIBV Eisenbahninfrastruktur-Benutzungsverordnung (DE) EIU Eisenbahninfrastrukturunternehmen (DE)

EPSF Etablissement public de sécurité ferroviaire (FR)

EU European Union

EVU Eisenbahnverkehrsunternehmen (DE) FOC Freight Train Operator (UK)

(18)

FOT Federal Office of Transport (SWI) FTE Full Time Equivalent

GDP Gross Domestic Product

GWF Greater Western Franchise (UK) HHI Herfindahl-Hirschman Index

HLOS High Level Output Specification (UK) HSE Health and Safety Executive (UK) HSL High Speed Line

ICA Italian Competition Authority IM Infrastructure Manager

IVW Inspectie Verkeer en Waterstraat (NL) JPIP Joint Performance Improvement Plan (UK) KKV Konkurrensverket (SE)

KPI Key Performance Indicators

KpVV Knowledge Centre on Public Transport (NL) LOTI Loi Orientation de Transport Intérieur (FR) LuFV Leistungs- und Finanzierungsvereinbarung (DE) MCAF Mission du contrôle des activités ferroviaires (FR) NEG National Express Group plc (UK)

Nma Nederlandse Mededingingsautoriteit (NL) NS Nederlandse Spoorwegen (NL)

OCCR Operationeel Controle Centrum Rail (NL)

OECD Organisation for Economic Co-operation and Development OFT Office of Fair Trading (UK)

OPRAF Office of Passenger Rail Franchising (UK) ORR Office of Rail Regulation (UK)

ORTF Organisation et Régulation des Transports Ferroviaires (FR) PI Performance Indicators

PPM Public Performance Measure (UK) PPP Public Private Partnership

PSO Public Service Obligation PTA Public Transport Authority

(19)

R&D Research&Development

RACO Railways Arbitration Commission (SWI) RegB Regierungskomission Bahn (DE)

RegTP Regulierungsbehörde für Telekommunikation und Post (DE) RFF Réseau Ferré de France (FR)

RMMS Rail Market Monitoring Scheme ROSCO ROlling Stock leasing Company (UK)

RU Railway Undertaking

RUS Route Utilisation Strategy (UK)

SBB-CFF-FFS

Swiss Federal Railways (SWI)

SKE- CACF- CAF-RACO

Schiedskommission im Eisenbahnverkehr, Commission d’arbitrage dans le domaine des chemins de fer, Commissione d’arbitrato in materia ferroviaria, Swiss Rail Arbitration Commission (SWI)

SNCF Société Nationale des Chemins de fer français (FR)

SOB Südostbahn (SWI)

SoFA Statement of Funds Available (UK) SRA Strategic Rail Authority (UK)

STIF Syndicat des transports d'Île-de-France (FR) TER Transport Express Regional (FR)

TET Trains d’Équilibre du Territoire (FR) TGN Thameslink and Great Northern (UK) TOC Train Operating Company (UK)

UIC Union Internationale des Chemins de fer

VDV Verband Deutscher Verkehrsunternehmen (DE)

VÖV UTP Verbandes öffentlicher Verkehr Union des transports publics, société cooperative Unione dei trasporti pubblici, società cooperativa (SWI)

VwVG Verwaltungsverfahrensgesetz (SWI) WEKO Wettbewerbskommission (SWI)

(20)

(21)

Introduction

This research project is about the liberalizing railway industry in Europe, particularly about the institutional dimensions of such liberalization. More precisely, we want to know whether liberalization – as an institutional change accompanied, among others, by the creation of competition and independent Regulatory Authorities – does lead to better performance of the railway systems. Our driving question is whether a causal relationship can be established between a given institutional arrangement and a certain performance of a railway system. In this introductory chapter we will briefly recall the context of this research project, formulate its main research questions, recall the theoretical framework and argumentation underlying this research, present its methodology, and outline the report’s broad structure.

1.1 The context and subject matter

In the European Union (EU) as of January 1st 2010 European rail undertakings (RUs) are granted a right of access to the rail infrastructure of other Member States for the purpose of operating international passenger services, including cabotage (Directive 2007/58/EC). Such market opening comes on top of the already existing access in the cargo market as of 2004 (Directive 2004/51/EC). While offering new business opportunities to RUs and infrastructures managers, the mandated access takes place in a still developing institutional environment. Regulatory Authorities in the EU, as well as the Swiss Parliament, Government, and administration, are still struggling with the application of this and the previous Directives (Directives 91/440/EEC amended by Directive 2004/51/EC and by Directive 2007/58/EC). For instance, Member States are allowed to limit the right of access on routes covered by public service contracts if certain conditions are met or to charge a levy on international rail passenger services to compensate for the costs incurred by such public service contracts. In the case of Switzerland, EU railway packages 1-3 have not yet been transposed into national legislation, even though freight transport is liberalized since 1999.

As witnessed by the opening of the European railway freight sector in 2007, the introduction of competition brought major institutional changes and required significant institutional adjustments. For example, most of the EU Member States have unbundled their historical railway operator (as mandated by Directive 91/440/EEC) and created some sort of Regulatory Authority. However, and despite the passing of the EU Directives that are common to all Member States, the institutional arrangements governing the railway sectors differ markedly from country to country. And so does the performance of the national railway sector, as well as the overall benefits derived from liberalization.

An increasing body of literature covers the introduction of competition in the freight sector at the European level (Bozicnik, 2009; Brewer, 1996; Pittman, 2005) and the introduction of competition in the passenger sector at the domestic level (UK, Germany). However, there is still little research and literature about the institutional arrangements that accompany such market opening, especially when it comes to competition in the international passenger market as well as to the broader opening of the domestic passenger market.

This research project aims at identifying, describing and analysing the main types of institutional arrangements emerged in a selected number of countries as a result of freight liberalization and now also international passenger liberalization. It attempts to draw a typology from the institutional landscapes observed and relate the arrangements to their

(22)

performance. By institutional arrangement we mean – in the intellectual tradition of institutional economics – the different actors involved in governing the railway sector of a country (e.g., RUs, infrastructure companies, sector specific Regulators, competition Regulators, price Regulators, national offices, political actors, legal actors and others more), their responsibilities (decision rights), as well as the formal and informal rules governing the relationships among these actors (distribution of these decision rights among these actors).

1.2 The research questions

Within the above outlined context, this research seeks to establish a relationship between a particular institutional arrangement on the one hand and the performance of a national railway system under this very institutional arrangement on the other. More precisely, we try to answer the following three questions:

- What are the various institutional arrangements in the European Member States and Switzerland as answers to the opening of the freight, as well as the international and in some cases the domestic passenger railway markets? Particular attention will be paid to the roles and responsibilities of the newly created Regulatory Authorities (where they exist), as well as to the roles and responsibilities of the Competition Authorities.

- How do these different institutional arrangements relate to the performance of the respective national railway systems? In other words, which institutional arrangement proves to be the most successful?

- What lessons can be learned from these different institutional arrangements and their performance?

The following Table 1 provides an overview of the parts of the report that are directly relevant to each research question.

Table 1. Research questions and sections of the report addressing them

What are the various institutional arrangements in the European Member States and Switzerland as answers to the opening of the freight, as well as the international and in some cases the domestic passenger railway markets?

- Review in chapter 3

- Detailed representation of actors’ evolution against time on the attached spreadsheets (referred to in chapter 3)

- Milestones of actors’ evolution against time on the attached spreadsheets (referred to in chapter 3) How do these different institutional

arrangements relate to the performance of the respective national railway systems?

- Review of performance measures in chapter 4 - Discussion of performance measures and

institutional developments in chapter 4

- Interactive charts of institutional arrangements and performance indicators on the attached spreadsheets (referred to in chapter 3)

What lessons can be learned from these different institutional arrangements and their performance?

- Conclusions on the institutional arrangements in chapter 3

- Interviews and conclusions on the role of the actors and the governance of the system in chapter 5

(23)

1.3 The underlying theoretical framework and argumentation

Let us briefly outline here the theoretical framework underlying this research project. The so-called “coherence framework” (Finger, et al. 2005) has been developed for the network industries in general but is applied here to the particular case of railways. It constitutes a way of conceptualizing the co-evolution between technology and institutions in the network industries, relating this co-evolution to the network industries’ performance. At its most general level, the framework states that there must be “some sort of” coherence between the state of technology in the network industries and the institutional arrangements governing this state of technology. If this coherence is “insufficient”, there will be consequences for the performance of the network industries, i.e., of the respective infrastructure system (e.g., air transport system, railway system, electricity system). Furthermore, the framework identifies a certain number of so-called “critical, system-relevant functions”, which are particularly sensitive to a “lack of coherence”. More precisely, four such critical, system-relevant functions are identified, namely interconnection, interoperability, capacity management and system management.

Historically, the network industries are characterized above all by a change in the institutional framework, such as liberalization, deregulation and privatization. This, it is argued, creates a certain “incoherence” with the existing state of technology in a given network industry, thus affecting performance. Telecommunications industry, where the initial change was technological in nature, may be an exception, because the change in institutions (e.g., liberalization) constitutes actually a step towards recreating certain coherence.

This general framework is now applied to the specific case of national railway systems: liberalization, deregulation and privatization (in the case of the UK) are said to introduce certain incoherence in the way the critical, system-relevant functions were governed historically, thus affecting the railway system’s overall performance. The assumption is that these critical, system-relevant functions must nevertheless be assumed. But the problematic issues of who takes these responsibilities and how to coordinate the different actors now partially responsible for these functions have emerged. Basically, the question is how the coordination (or lack thereof) among the actors in a (partially) liberalized railway environment relates to the performance of a given national railway system.

The (partial) liberalization of the railway sector, as for example promoted by the European Commission, leads to a new conceptualization of its governance, whose main characteristics are as follows:

- There is an infrastructure organization operating the tracks and the stations, unbundled at least in accounting terms from the train operating companies (or the rolling stock companies for that matter); as a sub-position, one may also separate infrastructure from stations.

- There is an independent Regulatory Authority that supervises the infrastructure operator (in the interest of the consumers); the Regulator also watches out on discrimination regarding the access to and pricing of the infrastructure.

- The question of financing is a matter of public policy and leads to different incentives depending on where and how financing occurs. The intellectually clean position is that there is (1) financing for the infrastructure operator and (2) financing for particular lines under public service mandates after a tendering procedure.

(24)

We do not know whether such an idealized governance mechanisms actually works, meaning that we do not really know what the performance of such a system actually would be. Thus the need for research. However, we do know about the current performance of the (sometimes partially unbundled) railway systems, none of which is identical to this idealized governance mechanism. It is legitimate to claim that any future system – including a perfectly unbundled system – should be at least as efficient as the current system.

In order to determine the performance of any railway system, we have developed the following “storyline”. By storyline we mean the logic of how to approach and ultimately measure the performance of a railway system. The storyline goes as follows:

- Unbundling is an institutional change. However, railways are first of all a technological, and not an institutional system. Institutions are built to make the (technological) railway system function. At any given state of (railway) technology, institutions can help make the railway system function more or less well. The functioning of the technological (railway) system will depend, among others, upon a series of critical functions that must be performed. These are interoperability, interconnection, capacity management, and system management. These functions are always ensured by way of a combination of technology and institutional rules. In a fully integrated railway system, the rules for these critical functions are internalized, i.e., defined inside the integrated company. Once the system is totally unbundled, the rules for ensuring the critical functions of interoperability, capacity management and system management must be defined and assumed by some other institution (for example the infrastructure company or the Regulator). At the beginning of liberalization (unbundling), technology, however, has not yet changed. The question now is whether the overall railway system is still performing as well as at the time when the rules were internalized (or worse or better).

- Indeed, the performance of a (railway) system can be defined as the result of a complex interplay/interaction between a given state of technology and the rules that govern this technology, in particular the rules about the critical functions. More precisely, the state of the institutions and the state of technology combined constitute incentives for the involved actors, and it is the interplay between these actors (who are or can be inside and outside the integrated company) to ultimately produce the outcomes/performance of the overall (railway) system.

- One should be able to describe the performance of any given (railway) system at any given moment of time. We propose, in the case of infrastructure systems such as railways, operational, technical, economic, social, and environmental performance measures (see chapter 4). However, relating such performance (different types of performance) to the interplay between any given state of technology and any given state of institutions is difficult, and attributing particular performances to particular actors (given that we are dealing with a complex socio-technical system) is even more difficult. Theory, for example, predicts that the introduction of competition (by way of unbundling) leads to better (short term) economic performance. This may well be the case (even though this should be ultimately established on the basis of facts and not theory), but it is for example not clear whether such economic efficiency gains thanks to unbundling and competition are offset by losses in other performance measures (e.g., punctuality, accidents, incidents). Ultimately, it falls of course upon the political authorities – i.e., the ones who finance and subsidize the railway system – to determine which PIs they value more than others and to prioritize them.

- We thus proposed, at the outset of this project, to study which type of arrangement between technology and institutions (“model”) produces which performances (along the

(25)

different categories of performance). At this point, we propose to establish a purely empirical link between the different models and their performances. Rather than to derive the performance from the model (i.e., the particular type of arrangement between a state of technology and a state of institutions), we propose to explain a given performance by the model, i.e., to find causes which are due to a particular model and which explain the particular performance of that model. Ideally, of course, we would also look at the dynamics, i.e., to understand how changes in the performance can be attributed to changes in the model (i.e., changes in the relationship between technology and institutions).

1.4 The underlying methodology

The research project rests largely on qualitative methodology.

The analysis of existing literature on the governance and competition in the railway sector, as well as a collection of information on the latest developments of institutional arrangements, has led to choosing five EU case studies, besides Switzerland. As part of the case studies, a pictorial representation of the evolution of the railway sector for the case countries has been developed. The main discontinuities in institutional arrangements have then been identified as milestones to be used in the quantitative section of the work.

Quantitative indicators available through international institutions (Eurostat, OECD, European Energy Agency), and therefore comparable across countries, have been used as the basis for investigating the relationship and the possible existence of an empirical link between performance and evolution of institutions.

Finally a questionnaire was proposed to stakeholders from the case countries to investigate their view of the systemic approach to railways within their current institutional arrangement and their expectations on performance for the system and for their own organisation.

1.5 Outline of the report’s structure

Besides this introductory chapter, this report is structured into four main parts.

A first part, comprising chapter 2 and chapter 3, gives a brief summary of the literature on competition in railways and develops the case studies on which the research is based. The existing actors are described and the evolution of the stakeholders since 1988 is outlined for each case study, characterising the most interesting elements and developing the support for a quantitative analysis in the next main part as well as a typology of institutions. The following part is made up by Chapter 4. This is devoted to performance measures, a review of those available and the discussion of how they changed against the institutional evolution in the case studies. The views of the stakeholders on the systems in which they work and on the relevant performances have been collected in the next part, contained in chapter 5. While intermediate conclusions are given at the end of each of the previous main parts, the key conclusions are presented in the closing chapter 6.

(26)

(27)

2 Competition in railways: a literature review

2.1 Introduction

This chapter is divided into two sections. We start by reviewing the literature on competition in the railway sector. Subsequently, we analyse the current state of competition in the European railway sector. This will lead us, in the conclusion, to select the countries of our study.

2.2 State of the literature on competition in railways

This section proposes a brief literature review of competition and of some of the issues raised by the introduction of competition in the railway sector. In particular, it highlights the importance of defining precisely what is meant by competition and the question of the means and ends of competition.

In his discussion paper on the competition for long-distance passenger rail service, Preston (2009, page 4) notes that “rail competition, where it occurs, is likely to be small group in

nature. Market demand is often too thin to support a large number of operators, whilst there may be some economies of scale and density that limit the optimum number of firms in rail

markets. The relevant industry structure is therefore that of oligopoly competition”.1

While most economists and policy makers agree that competition should be introduced in the railway industry, agreement on how this should be done is lacking. According to the OECD (2006) the three different modes of competition2 in the rail sector are:

- Competition in-the-market between vertically-integrated rail companies. This form of competition requires the existence of at least two separate rail infrastructures capable of providing substitute rail services (e.g., two different rail routes between a given city-pair); this is the predominant form of competition in rail freight services in North America.

- Competition in-the-market between train operating companies with regulated access to track infrastructure (which may or may not be owned by one of the companies providing train services). This is the predominant form of competition in freight services in Europe and most of Australia. Pittman (2008) notes that competition in-the-market can be further broken down into a complete “vertical separation” model – the prohibition of the network operator from operating its own trains, a policy urged by DG Competition – and a “third party access model”, with a vertically integrated infrastructure and train company forced to allow access to its infrastructure to competing, non-integrated train

1_{In fact classical models assume competition occurs either in the price dimension (Bertrand competition) or in}

the service dimension (Cournot competition). The conventional wisdom is that where capacity is difficult to increase (e.g., rail) competition will be of the Cournot type, but where capacity can easily be increased (e.g., bus) competition will be of the Bertrand type (Quinet & Vickerman, 2004).

2_{Holvad et al. (2003) distinguish two further types of competition: capital market competition and product}

market competition. Capital market competition concerns rail company ownership models and covers the spectrum ranging from Government departments under direct political control to arm’s length Agencies and finally to private corporations. Product market competition concerns the scale at which rail services can be provided by different train operating companies and ranges from pure monopoly to perfect competition.

(28)

operating companies. These models tend to blur a bit, as intermediate solutions such as “accounting separation” are accepted as means of preserving ownership integration while making third party access terms more transparent, and thus (it is hoped) preventing discrimination. Although route competition has remained a feature in countries such as Japan, in-the-market competition between passenger rail operators has been limited. For-the-market competition, in the form of competitive tendering and franchising is more common in the passenger rail industry than in-the-market.

- Competition for-the-market between rail companies (either for integrated track-plus-train services or just for train services alone, operating under a regime of regulated access to the track infrastructure). This is the predominant form of competition for regional passenger services in many EU countries.

Overall, the European style of competition varies notably from competition among vertically integrated train and infrastructure enterprises, a model chosen by policymakers in the geographically large, freight-dominated countries of the Americas – originally by the United States and Canada, more recently by Mexico, Brazil, and Argentina3.

In general, the degree to which complete separation of train and infrastructure operations in Europe has actually taken place – as well as the degree to which actual competition among train operating companies has appeared – varies significantly by country. “Third party access” tends to be observed in countries that have nominally chosen the vertical separation model but have moved only part way toward achieving it – i.e., countries that have taken steps to open up the train sector to competition but have not fully separated the incumbent freight operator from its infrastructure operations, Germany being the most salient example of this, as it has instituted “accounting separation” but not “ownership separation” of trains and tracks.

Biztan (2003) examined the cost implications of competition over existing US freight rail lines by testing for the condition of cost subadditivity. He finds that: (1) there are economies associated with vertically integrated roadway maintenance and transport, suggesting that separating the two would result in increased resource costs; and (2) railroads are natural monopolies in providing transport services over their own network, suggesting that multiple-firm competition over such a network would result in increased resource costs. These findings suggest that policies introducing rail competition through “open access” or on bottleneck segments would not be beneficial from a cost perspective. Moreover, the price decreases necessary for the introduction of such competition to be beneficial would be large. In an early study of the opening of Sweden to internal competition Jensen (1998) finds that external competitive pressure is strong in most supply segments and that, focusing on loss of scale advantages, the transformation will result in significant costs. Comparing the potential for gains by competition against the costs, he concludes that increased efficiency by internal competition seems possible only for two train products: domestic combined transport and dedicated trains (both freight services).

Looking at the freight sector Pittman (2005) identifies three attributes that have proven problematic for recent experiments with vertical separation: a) a relatively high share of

3_{The “American model” can be divided into a “North American model” – the US and Canada, with an emphasis}

on origin-destination competition between “parallel” vertically integrated railways – and a “Latin American model” – Mexico, Brazil, Chile, and Argentina, with an emphasis on competition for the business of shippers and customers at particular points served by more than one railway.

(29)

network costs in total delivered service costs, b) an apparent persistence of economies of scale at the “competitive” train operations level, and, perhaps most importantly, c) strong economies of vertical integration that are focused on the interface point of wheel and rail – that is, exactly where vertical separation takes place4.

Mulder et al. (2005) look at two questions: (1) whether the vertical separation between network and operation caused a loss of economies of scope and to which extent other institutional changes have improved the coordination between infra manager and railway operators, and (2) which efficiency improvements in railway operations have occurred and to which extent they are related to the institutional changes. In general, the literature on vertical economies (i.e., economies of scope between train and infrastructure operations) is not consistent, but mild economies are at least suggested (C. Growitsch & Wetzel, 2006; Ivaldi & Mccullough, 2008).

Nash (2008) looks at the objectives of European railway reforms, the different models adopted for restructuring, the key elements of separation of infrastructure from operations, as well as competition both through open access and franchising and regulation and infrastructure charging. He concludes that separation of infrastructure from operations involves costs, but it constitutes the most effective way of achieving within mode competition. Where operations do not greatly overlap and open access passenger and freight are unimportant, leasing infrastructure to passenger franchisees may be effective, but the model of vertical integration as separate subsidiaries within a holding company structure makes it difficult to ensure a level playing field for new entrants and is only effective where the vertically integrated operator remains dominant.

Overall, the literature is not conclusive as to the optimal institutional model: which is the model that is most conducive to competition? And which form of competition and related instuitutional model is actually leading to the best performance? The literature is mostly silent about this latter point. More recently, Mizutani and Uranishi (2011) conducted an econometric analysis of railways in OECD countries and concluded that vertical separation is linked to cost reductions in the case of low service density and to cost rises with higher density.

State of competition in European railways

As pointed out by Nash and Matthews (2009), progress on the 2001 White Paper has so far been largely restricted to intra-modal competition. Strong competition has emerged in the freight business particularly on the crucial North-South axis through the Alps, even if concerns on institutional arrangements that in many Member States do not ensure fair competition between former national railways and new entrants remain. In their study on the introduction of competition for local passenger railway markets in the German State of Baden-Württemberg, Lalive and Schmutzler (2008) find that the competitively-procured lines enjoyed a stronger growth of the frequency of service than those that were not procured competitively, even after controlling for various line characteristics that might have had an independent influence on the frequency of service. Their results further suggest that the effects of competition may depend strongly upon the operator.

4

The third factor in particular is a clear illustration of the rationales discussed by Coase & Williamson for the broad vertical scope of a single firm and the disadvantages of relying on market transactions under certain conditions.

(30)

In its recent communication (European Commission, 2010), the Commission acknowledges that market access remains difficult for new entrants. In order to maintain its policy of railway revitalization, the Commission aims to (1) improve non-discriminatory access to service facilities, (2) enhance transparency of the railway market’s institutional framework, (3) enhance cooperation and coordination to facilitate international rail transport, (4) provide effective incentives for sound and sustainable financing, and (5) enhance regulatory body independence and competencies. It also addresses the issue of competence regarding non-discrimination (e.g., Regulatory or Competition Authority)5.

Giannino (2010) argues that “incumbent railway undertakings still have a relevant market power” and, when vertically integrated, “they have an incentive to exercise their power to frustrate the entry of new competitors into the market. Such practices may fall within the competition rules that prohibit abuse of dominant position”. The investigations conducted by the Italian Competition Authority (ICA) have shown the potential of discriminatory practices carried out by the incumbent (e.g., the case Rail Traction Company/Rete Ferroviaria

italiana-Ferrovie dello Stato, Nuovo Trasporto Viaggiatori or FS/GVG).

Table 2 illustrates the degree of competition in the European railway sector as measured by relative market share.

Table 2. Market opening in Europe (2008) - market shares

Country Freight Passenger

Incumbent Non-incumbent MO score Incumbent Non-incumbent MO score AT 86 14 .740 88 12 .774 BE 93.9 6.1 .882 100 0 BG 85.68 14.32 100 0 DE 78 22 .608 89.9 10.1 .792 DK - - 91 9 .828 EE 51 49 .0389 42.3 57.7 .179 EL 100 0 100 0 1 ES 95 5 .903 100 0 1 FI 100 0 1.00 100 0 1 FR 95 5 .810 100 0 1 HU 85.6 14.4 .733 98.2 1.8 .964 IE 100 0 100 0 1 IT - - - - LT 100 0 1 100 0 1 LV 90.43 9.57 .818 90.92 9.08 .824 NL 67 33 98 2 PL 76.03 23.97 .470 88.89 11.11 .790 PT - - - - RO 59.01 40.99 .350 98.9 1 .978 SE - - - - SI 100 0 1 100 0 1 SK 97.97 2.03 99.97 0.03 1 UK 0 100 .311 0 100 0.001 NO 79 21 .620 88 12 .770

Note: Based on RMMS questionnaire; MO score computed using HHI – 2008

Comment: For freight, some major European countries (e.g., Italy, Sweden, Denmark) are missing Source: Everis (2010)

5_{In the case of access to service facilities, the Commission recommends to extend the scope of Regulatory}

(31)

Table 3 below summarizes the current legal state of competition in the different market segments as defined by EU legislation.

Table 3. Legal state of competition in Europe

Type of traffic Legislation Effective opening

to competition Type of competition Comment Passenger – international 3rd packet, Directive

2007/58/EC 01/01/2010 In the market

Possible competition for cross-border regional traffic Passenger – cabotage 3rd packet, Directive

2007/58/EC 01/01/2010 In the market

Restriction if endangers PSO equilibrium Passenger – domestic long distance Proposition from

EC None For the market Refused by Council

Passenger – domestic regional

PSO regulation

1370/2077/CE Latest 2019 For the market

Protection of existing contracts – 10 year transition period

Source: Compiled by authors, adapted from Desmaris, 2010

2.3 Conclusion: selection of countries

The countries for our case studies have been chosen in such a way that they offer the most different views on the progressive change in institutions – particularly in terms of unbundling and regulation. As a start, we have used Nash’s (2008) typology as presented in Table 4 below.

Table 4. Three institutional models of railways in Europe Complete separation

(the Swedish model)

Holding company (the German model)

Separation of key powers (the French model) Britain Finland Denmark Lithuania Netherlands Norway Spain Portugal Slovakia Sweden Austria Belgium Germany Italy Latvia Poland Greece Czech Republic Estonia France Hungary Slovenia Luxembourg

Note: Ireland and Northern Ireland remain vertically integrated Source: CEC, 2006, and Nash, 2008

While Germany and France have been chosen to represent their category of restructuring, Britain, Sweden and the Netherlands have been chosen because they represent different forms of separation. Britain saw the most radical reform and offers the possibility of looking at regulatory institutions that have seen a long development. Sweden was the first European country to unbundle its State railways and is the first to fully open access on its rail infrastructure. Moreover, two of the case countries fully opened their network for passenger services (Germany and Sweden), two chose to control access by way of concessions or franchises (the Netherlands and Britain, respectively) while use of infrastructure in France is

(32)

reserved for the state railway company SNCF. The Netherlands are also particularly interesting as they have a multipolar, dense and densely used network with important transit freight traffic (port to hinterland traffic), which makes them to some extent comparable to the Swiss rail case. Given the purpose of this report it was important to include Switzerland in the analysis, which allows drawing conclusions.

Governance of competition in the Swiss and European railway sector